def check_groups(self):
for tag in self.group_tags:
# In gff, the attribute will have a lowercase first letter
gff_tag = tag[0].lower() + tag[1:]
if gff_tag in self.f.qualifiers:
for group in self.f.qualifiers[gff_tag]:
group = self.check_group(group)
if group != "":
if group not in self.groups:
self.groups.append(group)
else:
self.warnings.append(
GeneError(GeneError.GROUP_MULTIPLE_SAME, self,
{'group': group}))
if group not in self.allowed_groups:
self.errors.append(
GeneError(GeneError.GROUP_UNKNOWN, self,
{'group': group}))
else:
self.errors.append(
GeneError(GeneError.GROUP_UNKNOWN, self,
{'group': group}))
if len(self.groups) == 0:
self.groups.append('Unknown')
if not self.no_group:
self.errors.append(GeneError(GeneError.GROUP_NONE, self))
elif len(self.groups) > 1:
self.warnings.append(GeneError(GeneError.GROUP_MULTIPLE, self))
def check_deleted_name(self):
if 'Name' not in self.f.qualifiers or self.f.qualifiers['Name'][
0] == "" or self.f.qualifiers['Name'][0] == "true":
self.errors.append(GeneError(GeneError.DELETED_MISSING_NAME, self))
else:
self.name = self.f.qualifiers['Name'][0].strip()
if not re.match("^[A-Z]{2,3}[0-9]{5,8}-R[A-Z]$", self.name):
self.errors.append(
GeneError(GeneError.DELETED_WRONG_NAME, self,
{'name': self.name}))
def check_intron(self):
if len(self.f.sub_features) > 0:
exon_coords = {}
# Find positions
for mrna in self.f.sub_features:
for gchild in mrna.sub_features:
if gchild.type == "exon":
exon_coords[
gchild.location.start] = gchild.location.end
# Check minimum intron size
start_sorted = sorted(exon_coords)
previous_end = None
for exon_start in start_sorted:
if previous_end != None:
intron_size = exon_start - previous_end
if intron_size < 9:
self.warnings.append(
GeneError(
GeneError.INTRON_TOO_SMALL, self, {
'len': intron_size,
'start': exon_start,
'end': previous_end
}))
previous_end = exon_coords[exon_start]
def check_cds(self):
# Check the total length of CDS
if len(self.f.sub_features) > 0:
cdsLen = 0
for sub1 in self.f.sub_features:
for sub2 in sub1.sub_features:
if sub2.type == 'CDS':
start = sub2.location.start
end = sub2.location.end
if end > start:
cdsLen += end - start
else:
cdsLen += start - end
if cdsLen == 0:
self.errors.append(GeneError(GeneError.CDS_IS_NULL, self))
if cdsLen < 20:
self.warnings.append(
GeneError(GeneError.CDS_IS_SMALL, self, {'len': cdsLen}))
def post_validation(self):
# validate splitted and duplicated genes once we collected the whole list
for s in self.splitted_genes.keys():
if len(self.splitted_genes[s]) == 1:
for p in self.splitted_genes[s].keys():
gene = self.splitted_genes[s][p]
if len(gene.display_id) == 32 and re.match(
"^[A-F0-9]+$", gene.display_id
): # If there is no name, it's probably the cause of the problem
gene.errors.append(
GeneError(GeneError.PART_SINGLE, gene))
else: # If there is a symbol it's probably an incomplete gene
gene.warnings.append(
GeneError(GeneError.PART_SINGLE_NAMED, gene))
self.all_genes[gene.wa_id] = gene
for s in self.duplicated_genes.keys():
if len(self.duplicated_genes[s]) == 1:
for p in self.duplicated_genes[s].keys():
gene = self.duplicated_genes[s][p]
gene.errors.append(GeneError(GeneError.ALLELE_SINGLE,
gene))
self.all_genes[gene.wa_id] = gene
# Check symbol and name are unique
seen_symbols = {}
warned_symbols = []
for g in self.all_genes.values():
if not g.allele and not g.part and not g.is_deleted and g.symbol is not None:
if g.symbol not in seen_symbols:
seen_symbols[g.symbol] = g
else:
if g.symbol not in warned_symbols:
seen_symbols[g.symbol].errors.append(
GeneError(GeneError.SYMBOL_NOT_UNIQUE,
seen_symbols[g.symbol]))
warned_symbols.append(g.symbol)
g.errors.append(GeneError(GeneError.SYMBOL_NOT_UNIQUE, g))
seen_names = {}
warned_names = []
for g in self.all_genes.values():
if not g.allele and not g.part and not g.is_deleted and g.name is not None:
if g.name not in seen_names:
seen_names[g.name] = g
else:
if g.name not in warned_names:
seen_names[g.name].errors.append(
GeneError(GeneError.NAME_NOT_UNIQUE,
seen_names[g.name]))
warned_names.append(g.name)
g.errors.append(GeneError(GeneError.NAME_NOT_UNIQUE, g))
def check_symbol(self):
if 'symbol' not in self.f.qualifiers or self.f.qualifiers['symbol'][
0] == "" or self.f.qualifiers['symbol'][0] == "true":
self.errors.append(GeneError(GeneError.SYMBOL_MISSING, self))
else:
symbol = self.f.qualifiers['symbol'][0].strip()
self.display_id = symbol
if not re.match("^[A-Za-z0-9-_.()/]+$", symbol):
self.errors.append(
GeneError(GeneError.SYMBOL_INVALID, self,
{'symbol': symbol}))
elif re.match("^[A-Z]{2,3}[0-9]{5,8}-R[A-Z]$", symbol):
self.errors.append(
GeneError(GeneError.SYMBOL_NOT_ID, self,
{'symbol': symbol}))
else:
self.symbol = self.f.qualifiers['symbol'][0].strip()
def check_dbxref(self):
if 'Dbxref' in self.f.qualifiers:
for dbxref in self.f.qualifiers['Dbxref']:
splitted_dbxref = dbxref.split(":")
db = splitted_dbxref[0].strip()
for t in self.group_tags:
if t.lower() == db.lower():
self.errors.append(
GeneError(GeneError.GROUP_MISPLACED, self,
{'tag': t}))
if db.lower() not in ['go', 'pmid', 'ncbi', 'uniprot']:
self.warnings.append(
GeneError(GeneError.DBXREF_UNKNOWN, self,
{'dbxref': dbxref}))
if dbxref.startswith('GO'):
self.has_goid = True
def check_name(self):
if 'Name' not in self.f.qualifiers or self.f.qualifiers['Name'][
0] == "" or self.f.qualifiers['Name'][0] == "true":
self.errors.append(GeneError(GeneError.NAME_MISSING, self))
else:
name = self.f.qualifiers['Name'][0].strip()
if len(name) == 32 and re.match("^[A-F0-9]+$", name):
self.errors.append(
GeneError(GeneError.NAME_INVALID, self, {'name': name}))
elif 'putative' in name.lower():
self.warnings.append(
GeneError(GeneError.PUTATIVE, self, {'name': name}))
elif 'similar to' in name.lower():
self.errors.append(
GeneError(GeneError.SIMILAR_TO, self, {'name': name}))
elif '-like' in name.lower():
self.warnings.append(
GeneError(GeneError.SIMILAR_TO, self, {'name': name}))
elif re.match("^[A-Z]{2,3}[0-9]{5,8}-R[A-Z]$", name):
self.errors.append(
GeneError(GeneError.NAME_NOT_ID, self, {'name': name}))
else:
self.name = self.f.qualifiers['Name'][0].strip()
def get_tag_value(self, key, allowed=[]):
for qk in self.f.qualifiers.keys():
if key.lower() == qk.strip().lower():
new_value = self.f.qualifiers[qk][0].strip()
if len(allowed) > 0 and new_value not in allowed:
self.errors.append(
GeneError(GeneError.ATTRIBUTE_INVALID, self, {
'key': key,
'value': new_value
}))
return new_value
return None
def check_multiple_mrnas(self):
if len(self.f.sub_features) > 1:
gene_name = self.f.qualifiers['Name'][0]
for child in self.f.sub_features:
if child.type == "mRNA":
if len(
child.qualifiers['Name']
[0]) < len(gene_name) or not child.qualifiers['Name'][
0].startswith(gene_name) or not re.match(
"^ [A-F]{1,2}$",
child.qualifiers['Name'][0][len(gene_name):]):
self.errors.append(
GeneError(GeneError.INVALID_MRNA_NAME, self,
{'gene_name': gene_name}))
def validate_genes(self):
in_handle = open(self.in_file)
for rec in GFF.parse(in_handle):
for f in rec.features:
if (f.type == "gene") and (
'status' not in f.qualifiers
or not f.qualifiers['status']
or f.qualifiers['status'][0].lower() != "deleted"):
gene = Gene(f, rec.id, self.scaf_lengths[rec.id],
self.allowed_groups, self.group_tags,
self.no_group, self.split_users)
self.all_genes[gene.wa_id] = gene
# Count number of genes with goid
if gene.has_goid:
self.genes_with_goid += 1
# Collect stats on groups
for g in gene.groups:
if g not in self.groups_stats:
self.groups_stats[g] = 0
self.groups_stats[g] += 1
new_part = gene.part
new_allele = gene.allele
# Collect wa_errors
self.wa_errors.extend(gene.wa_errors)
if not new_part and not new_allele:
self.genes_seen_once += 1
# keep track of splitted genes
if new_part:
part_gene_key = gene.display_id
if new_allele:
part_gene_key = gene.display_id + ", allele " + new_allele
if part_gene_key not in self.splitted_genes:
self.splitted_genes[part_gene_key] = {}
if new_part in self.splitted_genes[part_gene_key]:
identical = self.splitted_genes[part_gene_key][
new_part]
gene.errors.append(
GeneError(
GeneError.PART_SAME, gene, {
'other_name': identical.display_id,
'other_scaff': identical.scaffold,
'other_start':
identical.f.location.start,
'other_end': identical.f.location.end
}))
self.splitted_genes[part_gene_key][new_part] = gene
# keep track of duplicated genes
if new_allele:
allele_gene_key = gene.display_id
if allele_gene_key not in self.duplicated_genes:
self.duplicated_genes[allele_gene_key] = {}
if new_allele in self.duplicated_genes[
allele_gene_key]:
identical = self.duplicated_genes[allele_gene_key][
new_allele]
if identical.part == new_part:
gene.errors.append(
GeneError(
GeneError.ALLELE_SAME, gene, {
'other_name': identical.display_id,
'other_scaff': identical.scaffold,
'other_start':
identical.f.location.start,
'other_end':
identical.f.location.end
}))
self.duplicated_genes[allele_gene_key][
new_allele] = gene
elif 'status' in f.qualifiers and f.qualifiers[
'status'] and f.qualifiers['status'][0].lower(
) == "deleted":
gene = Gene(f, rec.id, self.scaf_lengths[rec.id],
self.allowed_groups, self.group_tags,
self.no_group, self.split_users)
self.all_genes[gene.wa_id] = gene
else:
fake_gene = Gene(f, rec.id, self.scaf_lengths[rec.id],
self.allowed_groups, self.group_tags)
self.wa_errors.append(
WAError(WAError.UNEXPECTED_FEATURE, fake_gene))
in_handle.close()
def check_status(self):
if (not self.apollo_1x) and ('status' not in self.f.qualifiers or
(self.f.qualifiers['status'][0].lower()
== "needs review")):
self.errors.append(GeneError(GeneError.NEEDS_REVIEW, self))